Formation and Possible Reactions of Organometallic Intermediates

The complex has a high affinity for alkyl halides (ATRP pathway) but also has sufficient .... The first one leads to reversible metal–carbon bond fo...
0 downloads 0 Views 676KB Size
Article pubs.acs.org/Organometallics

Formation and Possible Reactions of Organometallic Intermediates with Active Copper(I) Catalysts in ATRP Kristin Schröder,† Dominik Konkolewicz,† Rinaldo Poli,‡,§ and Krzysztof Matyjaszewski*,† †

Department of Chemistry, Carnegie Mellon University, 4400 Fifth Avenue, Pittsburgh, Pennsylvania 15213, United States Laboratoire de Chimie de Coordination (LCC), CNRS, Université de Toulouse, UPS, INPT, 205, Route de Narbonne, 31077 Toulouse, France § Institut Universitaire de France, 103, Boulevard Saint-Michel, 75005 Paris, France ‡

S Supporting Information *

ABSTRACT: The CuI complex obtained in situ from CuI and tris((4methoxy-3,5-dimethylpyridin-2-yl)methyl)amine (TPMA*) is currently the most reducing and the most active catalyst for atom transfer radical polymerizations (ATRP). The complex has a high affinity for alkyl halides (ATRP pathway) but also has sufficient affinity toward organic radicals to potentially participate in organometallic-mediated radical polymerization (OMRP). Thus, the radical polymerization of n-butyl acrylate initiated by AIBN (azobisisobutyronitrile) was significantly retarded, and the molecular weights decreased in the presence of the CuI/TPMA* complex. These results suggest the presence of a Cu-mediated termination processes, even after the amount of radicals generated from AIBN exceeded the initial amount of CuI/TPMA*. Nevertheless, in the presence of alkyl bromides, which act as ATRP initiators for acrylates, control was gained through metal-mediated halogen atom transfer, i.e., ATRP, not OMRP.



INTRODUCTION Conventional radical polymerization (RP) procedures are the most commonly employed industrial techniques for the production of commodity polymers.1 RP is characterized by good tolerance to functional groups and many impurities, but provides poor control over polymer architecture, molecular weight (Mn), and molecular weight distribution (Mw/Mn). In contrast, reversible-deactivation radical polymerizations (RDRP),2 also named controlled/living radical polymerizations (CRP), can provide control over Mn, Mw/Mn, and access to polymers with predefined functionalities, compositions, and architectures.3 Transition metal-mediated RDRP techniques include atom transfer radical polymerization (ATRP)4 and organometallic-mediated radical polymerization (OMRP).5 They rely on a dynamic equilibrium between a low fraction of growing radicals and a predominant fraction of dormant species. Since the pioneering work on OMRP by Wayland and coworkers, utilizing cobalt porphyrin complexes for the polymerization of acrylates,6 the procedure has been expanded to employ various complexes based on a range of transition metals, e.g., titanium, chromium, iron, cobalt, and others,7 for a large number of monomers, including vinyl acetate8 and acrylamides.9,10 OMRP utilizes a fast reversible homolytic cleavage of a transition metal−carbon bond, Scheme 1. Further OMRP reactions are often initiated by conventional organic radical sources, e.g., azobisisobutyronitrile (AIBN) in the presence of a transition metal complex in its lower oxidation state. The concentration of growing radical chains (Pn*) is © 2012 American Chemical Society

Scheme 1. Schematic Illustration of ATRP and OMRP Pathways

controlled by a formal reversible one-electron process, in which Pn* reacts with a redox-active transition metal complex (MtnL), resulting in the formation of an oxidized organometallic complex (Mtn+1−Pn/L) in the so-called reversible termination process (RT OMRP). The dynamic active/dormant species equilibrium, strongly favoring the dormant state, i.e., Mtn+1−Pn/ L, should ensure a low constant concentration of propagating radicals (Pn*), allowing concurrent growth of all polymer chains and minimizing bimolecular termination reactions. Additional control may be provided through a degenerative transfer process (DT OMRP), which involves the reversible associative exchange between Mtn+1−Pn/L and another Special Issue: Copper Organometallic Chemistry Received: July 22, 2012 Published: October 31, 2012 7994

dx.doi.org/10.1021/om3006883 | Organometallics 2012, 31, 7994−7999

Organometallics

Article

propagating radical (Pm*), releasing the previous dormant radical (Pn*).8a Consequently, in an OMRP all resulting dormant polymer chains are metal-capped; therefore OMRP requires the presence of stoichiometric amounts of MtnL. In some cases, catalytic chain transfer (CCT) reactions, Scheme 1, may be involved, especially when using cobalt as the transition metal for the polymerization of methacrylates.8b,11 Formally, β-hydrogen transfer (kCCT) between Pn* and MtnL yields an olefin-terminated species (P=) and a metal hydride complex, H−Mtn+1L. The H−Mtn+1L complex reinitiates polymerization of a vinyl monomer, forming MtnL and a propagating radical. In contrast to OMRP, ATRP gives halogen-capped polymers and is controlled by a fast and reversible homolytic cleavage of a C−halogen bond in a redox reaction with a transition metal catalyst, Scheme 1, with KATRP = ka,ATRP/kd,ATRP.12 Since ATRP is a true catalytic process, the transition metal complex can be used in substoichiometric concentrations. However, as is the case in OMRP and for the same reasons, the dynamic active/ dormant species equilibrium must strongly favor the stable dormant state, in this case Pn−X. Recent developments in ATRP,13 e.g., initiators for continuous activators regeneration (ICAR),14 target the in situ regeneration of the activator complex (Mn/L), from the oxidatively stable deactivator complex (Mn+1X/L), formed by unavoidable termination reactions. They allow the transition metal complex to be utilized at parts per million (ppm) concentrations. Values of KATRP can vary over a wide range.15 Simply altering the catalyst (MtnL) or initiator (R−X) structure provides a powerful tool to influence KATRP, in each case by more than 6 orders of magnitude.16 Recently, active catalysts for ATRP were developed by incorporating electron-donating groups into pyridine-based ligand structures such as 2,2′-bipyridine (bpy)17 and tris(2-pyridylmethyl)amine (TPMA) scaffolds.18 Tris((4-methoxy-3,5-dimethylpyridin-2-yl)methyl)amine (TPMA*) currently provides the most active ligand developed for Cu-catalyzed ATRP. Part of a quest to develop a deeper understanding of the mechanisms of transition metal mediated CRP has focused on the potential interplay between ATRP, OMRP, and CCT mechanisms.5a,7b ATRP and OMRP can interact synergistically (i.e., the two equilibria, each providing complementary control over the concentration of the propagating radicals) since the ATRP catalyst in the low oxidation state may also act as an OMRP radical trap. Examples of such systems include Momediated19 and Os-mediated20 polymerizations conducted with phosphine-containing complexes. However, if the equilibria are too strongly displaced toward the dormant species, then the one-electron oxidative addition between Mtn/L and R−X leading to Mtn+1−X/L and Mtn+1−R/L becomes irreversible and, essentially, no polymerization takes place. OMRP and CCT may competitively interplay, because both processes involve the same reagents Mtn/L and Pn*. The first one leads to reversible metal−carbon bond formation, and the second one to β-H atom transfer. The former process can also include β-H elimination from the OMRP dormant species Mtn+1−Pn/L to yield the same product as in CCT. Such reactions were reported for Mo-mediated polymerizations19a and Fe-mediated processes with α-diimine21 or aminebis(phenolate) ligands.22 For Cu-catalyzed ATRP processes, only one report, in 1998, pointed to the possibility of an interaction between an organic radical and a copper(I) complex in the absence of halogen

atoms.23 Diminished polymerization rates for methyl acrylate (MA) initiated by AIBN were observed in the presence of CuIOTf complexes with alkyl-substituted bpy ligands. The generation of sufficiently stabilized RCuII dormant species is notable, since stable alkyl derivatives of CuII are scarce.24 Furthermore, the possibility of a β-H atom transfer process, leading to CuII-H intermediates, has not yet been discussed. Herein, we investigated the role of Cu/TPMA* complexes under ATRP (with R−X and AIBN) and OMRP (only AIBN) conditions. After the initial evaluation of the effect of Cu/ TPMA* complexes on KATRP under ATRP conditions,18 an examination of the effect of the complex under OMRP conditions revealed a significant influence on polymerization rates and molecular weight distributions. The unprecedented contribution of Cu complexes to catalyzed radical terminations is subsequently critically evaluated.



RESULTS AND DISCUSSION KATRP and ka for Cu/TPMA*, Cu/dNbpy, and Cu/bpy. The values of KATRP and ka for the CuI/TPMA* catalyst complex were determined via stopped-flow measurements using methyl 2-bromopropionate (MBP) as an initiator in acetonitrile at 25 °C.16 A value for KATRP of 4.2 × 10−4 for Cu/ TPMA*18 was determined from the modified Fischer equation25 based on the persistent radical effect (PRE), cf. SI.26 The value of KATRP obtained for Cu/TPMA* is more than 5 orders of magnitude greater than that determined for Cu/ dNbpy and 6 orders of magnitude greater than that measured for the unsubstituted Cu/bpy catalyst complex, KATRP = 1.0 × 10−9 and KATRP = 1.3 × 10−10, respectively (Scheme 2).16 Scheme 2. Effect of Ligands on KATRP in Cu-Catalyzed ATRP Processes under the Same Conditions (MBP, MeCN, 25 °C)

An activation rate coefficient, ka,TPMA* = 8400 M−1 s−1, was determined by trapping the generated methyl 2-propionate radicals with 2,2,6,6-tetramethylpiperidinyl-1-oxy (TEMPO). This value, in comparison with reported values16 of ka,dNbpy = 3.7 × 10−2 M−1 s−1 and ka,bpy = 4.0 × 10−3 M−1 s−1, correlated well with the trend observed for KATRP. The high activation rate coefficient, combined with the high equilibrium constant, indicates that deactivation rate coefficient is sufficiently fast to provide good control in an ATRP (kd,TPMA* = 2 × 107 M−1 s−1). Normal ATRP and ICAR ATRP of Acrylates with Cu/ TPMA* in the Presence of an Alkyl Halide. First, “normal” and ICAR ATRP were investigated to assess the viability of copper complexes formed with TPMA* for the polymerization of acrylates. The high catalyst loadings in a “normal” ATRP, combined with the very high activity of CuI/TPMA*, initially created very high radical concentrations, which resulted in significant termination reactions and relatively poor control. On the contrary, ICAR ATRP, which relies on continuous regeneration of low concentrations of CuI/TPMA*, showed excellent control over Mn with concentrations of catalyst as low 7995

dx.doi.org/10.1021/om3006883 | Organometallics 2012, 31, 7994−7999

Organometallics

Article

conditions), the polymerization reached only ∼2% after 45.8 h. Assuming a quantitative reaction between CuI and radicals generated from AIBN ( f = 0.8) to yield an R−CuII OMRP dormant species, complete consumption of CuI should occur after 1.6, 3.2, and 6.8 h for systems with 100 ppm CuI (5% AIBN consumption), 200 ppm (10%), and 400 ppm (20% AIBN consumption), respectively. This can be determined from the exponential decay in the AIBN concentration over time, [AIBN]t = [AIBN]0 exp(−kazot). Setting the loss of AIBN equal to the initial [CuI] gives the time (t) needed to react all CuI, i.e., [CuI]0 = 2f([AIBN]0 − [AIBN]t) = 2f [AIBN]0 (1 − exp(−kazot)); t = −(1/kazo) ln(1 − ([CuI]0/2f [AIBN]0)). However, nearly linear pseudo-first-order kinetic plots were observed for more than 30 h, indicating a relatively constant amount of radicals in the system. This observation confirms that a radical quenching process must operate not only in the presence of CuI/TPMA* but also in the presence of the putative R-CuII/TPMA* species. The progressive decrease of the apparent propagation rate constants, kapp,100 ppm = 0.049 h−1, kapp,200 ppm = 0.017 h−1, kapp,400 ppm = 0.008 h−1, follows the amounts of CuIBF4/TPMA* (Table S2). In all cases there is a decrease in the rate that is proportional to the concentration of the CuIBF4/TPMA*, within the standard deviations of the 100, 200, and 400 ppm experiments. This suggests the possibility of radical trapping and/or a catalytic retardation process. It should be stressed that the kapp,100 ppm for CuIBF4/TPMA* was nearly 8 times smaller than kapp of the ICAR ATRP reactions using the same amount of Cu and AIBN but in the presence of alkyl bromides (SI Table S1, entry 5, kapp,100 ppm = 0.36 h−1). Bimodal gel permeation chromatography (GPC) traces for polymer prepared in the conventional radical polymerization reactions were observed (Figure 2A; SI, Table S3). They can be explained by an initial rapid exothermic polymerization and a two-stage process. First, a high molecular weight fraction could be quickly formed in a concentrated monomer solution (Mn = 8 × 105 to 1 × 106). Then, due to the polymerization exothermicity, the temperature could increase, accelerating AIBN decomposition and resulting in a lower molecular weight fraction (Mn ≈ 2 × 104). A temperature probe directly inserted into the reaction mixture showed a temperature increase by >50 °C during the polymerization (SI, Figure S5). Increasing the anisole content to 90% (v/v) slowed the polymerization and resulted in monomodal distribution, confirming this assumption (SI, Figure S4). The evolution of Mn and Mw/Mn with conversion in the presence of CuIBF4/TPMA* indicated a significant difference compared to conventional RP. The GPC traces (Figure 2A) of the polymers obtained in the reactions conducted in the presence of CuIBF4/TPMA* showed narrower molecular weight distributions (MWD); Mw/Mn ≈ 1.6 (SI, Figure S2B). Nevertheless, Mn,exp did not show a linear increase with conversion (SI, Figure S2B), and the Mn,exp values were similar to or lower than the Mn,theo values, calculated from the actual number of chains generated by the decay of AIBN (SI, Figure S3) or the amount of CuIBF4/TPMA*, respectively (Figure 2B). This could suggest the participation of a chain transfer process, although less efficient than in a typical CCT, which normally gives oligomers.19a Potential Polymerization Mechanisms. Since the radical concentration in the system with CuI is much lower than the conventional RP, a radical trapping by the CuIBF4/TPMA* complex is likely. Since an excess of TPMA* was used to ensure efficient formation of the CuIBF4/TPMA* complex, the ligand itself could also cause retardation. However, a blank experiment

as 100 ppm (Mw/Mn = 1.15) down to 5 ppm (Mw/Mn = 1.43), cf. SI Table S1 and Figure S1. These findings agree with the results of PREDICI simulations and indicate that very active ATRP complexes are better suited for low ppm ICAR ATRP, than normal ATRP.18 OMRP Initiated by AIBN in the Presence of CuIBF4/ TPMA* in the Absence of an Alkyl Halide. The very strongly reducing and active CuIBF4/TPMA* complexes could potentially directly interact with propagating radicals. First, control experiments with BA and AIBN under typical conventional radical polymerization conditions were performed and exhibited the expected uncontrolled behavior with very fast kinetics and broad molecular weight distributions (Mw/Mn ≈ 20) (Figures 1 and 2A, green).

Figure 1. Kinetic plots for polymerization with various loadings of [CuI]. Conditions: [BA]:[AIBN]:[TPMA*]:[CuI(CH3CN)4BF4] = [160]:[0.2]:[0.06−0.24]:[0.016−0.064], [BA] = 5.6 M, 20% (v/v) anisole, T = 60 °C.

Figure 2. (A) Normalized GPC traces in the presence of 0, 100, 200, and 400 ppm CuIBF4/TPMA* and (B) Mn,exp (dots) and Mn,theo (lines) as function of conversion. Parameters and calculation of values for Mn,theo(AIBN) and Mn,theo(CuIBF4/TPMA*) are provided in the SI. Conditions: [BA]:[AIBN]:[TPMA*]:[CuI(CH3CN)4BF4] = [160]:[0.2]:[0.06−0.24]:[0.016−0.064], [BA] = 5.6 M, 20% (v/v) anisole, T = 60 °C.

Then, a series of polymerizations of BA initiated by AIBN in the presence of various concentrations of the CuIBF4/TPMA* complex was carried out. The experiments with 100, 200, and 400 ppm CuI were repeated twice, and average values and standard deviation for the pseudo-first-order kinetic plots are presented in the SI, Table S2. In the presence of 100 ppm of CuIBF4/TPMA*, the rate of polymerization was much slower than in conventional radical polymerization. Conversion increased in a linear fashion in semilogarithmic coordinates and after nearly 30 h reached only 76%, Figure 1 and SI Figure S2A. Increasing the amount of CuI complex, to 200 ppm and then to 400 ppm (mol, vs monomer), resulted in even slower reactions, reaching after 31 h only 55% and 25% monomer conversion, respectively. With a [AIBN]:[CuI] ratio of 1:1, corresponding to 1250 ppm Cu (i.e., under typical OMRP 7996

dx.doi.org/10.1021/om3006883 | Organometallics 2012, 31, 7994−7999

Organometallics

Article

In case of rapid formation of the CuII intermediates, followed by slow termination with propagating radicals, a square root dependence of kapp on the [AIBN] would result. In the OMRP scenario, polymerizations in the presence of CuIBF4/TPMA* are slower, and, consequently, after longer reaction times, a greater number of radicals are generated from AIBN, leading to lower molecular weights. Organometallic copper(II) compounds are intrinsically unstable and therefore rarely studied,27 although they were postulated as intermediates in aqueous solutions.28 Decomposition of paramagnetic R− CuII/TPMA* (S = 1/2) might take place through termination with another radical (R*) to produce R−R by recombination and regenerate CuI/TPMA*. Radical−radical coupling processes were previously reported.7b,24b This pathway is in agreement with observed Mw/Mn ≈ 1.6 and constant low polymerization rates. Furthermore, decomposition of R−CuII/ L can occur via additional routes, i.e., (1) a one-electron reduction processes, yielding a CuI complex and a radical (OMRP reactivation), (2) heterolysis of the carbon−copper bond to give a CuII complex and saturated alkyl species (formed from the carbanion in the presence of H2O), (3) a reductive elimination processes (two-electron reduction) yielding Cu(0), or (4) β-elimination processes.28 The formation of Cu(0) could be supported by the appearance of small particles in the homogeneous polymerization mixture. Note, however, that Cu(0) can also result from disproportionation. Another possibility is the formation of an unsaturated alkene and the paramagnetic copper(II) hydride (S = 1/2), H−CuII/ TPMA*. The latter can terminate with another radical or reinitiate polymerization with another monomer. Reinitiation would start a CCT polymerization similar to that known for the cobalt/methacrylate systems.11 However, the formation of

with a [TPMA*]:[AIBN] ratio of 0.24:0.2 carried out under conditions similar to those with 400 ppm of CuIBF4/TPMA* exhibited a conventional RP behavior with 77% conversion after 0.5 h, forming a polymer with a broad MWD. In addition, a blank experiment utilizing CuI(CH3CN)4BF4 salt (without TPMA*) also resulted in conventional RP behavior (72% conversion after 0.5 h with Mw/Mn = 13.6). This indicates that only the CuIBF4/TPMA* complex interferes with the conventional radical chain growth process. Scheme 3 proposes two feasible pathways for the Cu-mediated reactions. Scheme 3. Possible Pathways during the AIBN-Initiated Polymerization of BA with CuIBF4/TPMA* (= CuI/TPMA*)

Polymerizations with different initial AIBN concentrations (SI, Figure S6) were performed in order to determine the ratedetermining step for both pathways. A linear dependence of kapp on the [AIBN] (SI, Figure S7) was observed. As outlined in the Supporting Information, this linear dependence of the radical concentration and consequently the polymerization rate is consistent with slow formation of the R−CuII intermediates (rate-determining step) followed by fast termination reactions between the R−CuII intermediates and the propagating radicals.

Figure 3. Superposition of simulated (lines) and experimental (circles) data for the polymerization of BA at various [CuI] loadings: green, 0 ppm (conventional radical polymerization); blue, 100 ppm; red, 200 ppm; black, 400 ppm; brown, 1250 ppm. (A−C) OMRP with deactivation and termination (kd,OMRP = 1.6 × 105 M−1 s−1, kt,OMRP = 3 × 108 M−1 s−1): (A) Kinetic plots, (B) DPn and (C) Mw/Mn vs conversion. (D−F) Catalytic radical termination (CRT) (kd,CRT = 1.6 × 105 M−1 s−1, kt,CRT = 3 × 108 M−1 s−1): (A) Kinetic plots (B), DPn and (C) Mw/Mn vs conversion. Conditions: [BA]:[AIBN]:[CuI/TPMA*] = [160]:[0.2]:[0.016−0.2], [BA] = 5.6 M, T = 60 °C. Points refer to experimental data; lines refer to kinetic simulations. 7997

dx.doi.org/10.1021/om3006883 | Organometallics 2012, 31, 7994−7999

Organometallics

Article

or controllability. In contrast, in the absence of RBr, only the OMRP or CRT pathways are possible and the rate of polymerization is decreased, as well as the control over the polymerization. This difference is due to the fact that the ATRP equilibrium shifts nearly all CuI species to X−CuII, effectively removing the majority of the CuI complex that can terminate propagating chains from the reaction medium. However, due to the high activity of the TPMA*-based catalyst complexes, the minute amounts of CuI are still very active in ATRP, and the low mole fractions are sufficient for the continuous reactivation of dormant species. Comparison with Previously Observed Interactions with Cu/dTbpy and Cu/dNbpy. The AIBN-initiated polymerization of MA in the presence of CuIOTf/dTbpy reduced the polymerization rates.23 While the polymerization conditions used in the present work were similar to those of the previous report (for a detailed comparison, see SI Table S4), there was a much larger decrease in the polymerization rate for a polymerization of BA (kp, 60 °C = 33 700 L mol−1 s−1)29 conducted in the presence of Cu I BF 4 /TPMA* (8.5% conversion after 3 h) than for the polymerization of MA (kp, 60 °C = 33 100 L mol−1s−1)29 in the presence of CuIOTf/ dTbpy (∼35% conversion after 3 h), although solvent effects can also have some contributions.15d There were also significant differences in the resulting Mn and Mw/Mn values. While the Mn values for the PMA polymers formed in the presence of the CuIOTf/dTbpy complex were ∼2 × 105, the PBA polymers obtained in the presence of CuIBF4/TPMA* had Mn values ∼3 × 104. These results suggest that the substantially more reducing and more active CuIBF4/TPMA* ATRP catalyst interacts with the growing acrylate radicals stronger than the less active CuIOTf/dTbpy complex.

polymers with relatively high molecular weights suggests that the copper hydride intermediate undergoes catalytic radical termination (CRT) with another radical, rather than reinitiation. PREDICI Simulations. PREDICI simulations were performed for the two most likely scenarios in order to get further insights into the potential polymerization mechanisms, cf. SI for conditions. Firstly, the OMRP process was considered. Note that, with only deactivation taking place, OMRP shows no sustained retardation; the simulations are supplied in the SI (Figure S8). OMRP including deactivation and termination was simulated (kd,OMRP = 1.6 × 105 M−1 s−1, kt,OMRP = 3 × 108 M−1 s−1, Figure 3A−C), and the observed retardation agrees well with the experimental data. Furthermore, polymerizations were slow and depended on the amount of CuIBF4/TPMA* added to the reaction, while the Mw/Mn values were ∼1.5. Similarly, CRT (kd,CRT = 1.6 × 105 M−1 s−1, kt,CRT = 3 × 108 M−1 s−1) was considered (Figure 3D, E). The presence of transfer reactions did not alter the simulated kinetics, and molecular weights were in agreement with the experimentally observed values. However, the Mw/Mn values did not decrease below ∼2. Note that the simulations in Figure 3D−F are indistinguishable from a pathway resulting in an initial formation of the R−CuII/ TPMA* organometallic species. This could be followed by rapid β-elimination to give H−CuII/TPMA* and an unsaturated chain and finally a rapid termination between radicals and the H−CuII/TPMA* to give a dead chain and CuI/TPMA* (SI, Figure S9), assuming kd,OMRP = 1.6 × 105 M−1 s−1, kβ = 3 × 108 s−1, and kt,CRT = 3 × 108 M−1 s−1. The concentrations of the various Cu species under the OMRP or the CRT conditions are given in the SI (Figure S10). In all cases, CuI concentrations are constant throughout the reaction, and [CuII] is very low (∼1 μM), since the reaction between CuI and the propagating radical is the rate-limiting step. Consequences for ATRP. Under the same conditions (temperature, [AIBN], [Cu]), the polymerization is much faster in the presence of RBr (ICAR ATRP). With 100 ppm CuIBF4/TPMA*, the apparent rate constant is kapp,100 ppm = 0.048 h−1, and in the presence of 100 ppm CuIICl2/TPMA* (and 0.037 M alkyl bromide) the apparent rate constant is kapp,100 ppm = 0.36 h−1. Therefore, the contribution of interactions of radicals with CuI/TPMA* (retardation) to the overall ICAR ATRP rate is not very significant. Consequently, one can conclude that the control observed under ATRP conditions can be assigned to the ATRP equilibrium. ICAR ATRP of BA with 100 ppm catalyst loading showed excellent agreement between Mn,exp and Mn,theo, providing polymers with low values Mw/Mn = 1.15, whereas the polymerization in the presence of CuIBF4/TPMA* in the absence of an ATRP initiator showed uncontrolled characteristics, generating polymers with Mw/Mn = 1.88 after 4 h. Indeed PREDICI simulations show that under ATRP conditions the polymerization of BA with AIBN at 60 °C is much better controlled in the presence of RBr (SI, Figure S11). The modeled system agrees well with the results of reactions performed earlier for butyl acrylate. These simulations showed that in the presence of alkyl halides the rate of polymerization closely matched that of the RP process, regardless of whether the system was initiated in the presence of CuI or CuII. In both cases control over the molecular weight was very good. Furthermore, inclusion of either OMRP, followed by termination of the organometallic species, or CRT of the CuII hydride into this ATRP scheme resulted in minimal impact in terms of the polymerization rate



CONCLUSIONS



ASSOCIATED CONTENT

Interactions between radicals and highly active ATRP complexes (CuIBF4/TPMA*) were observed in the polymerization of acrylates. These interactions led to significantly slower polymerizations, lower molecular weights, and narrower molecular weight distributions. This suggests that OMRP trapping processes contribute to retarding the polymerization and preventing the initial formation of polymers with very high molecular weights. Furthermore, the experiments described here also demonstrate the existence of additional termination processes, in the presence of the CuIBF4/TPMA* complex, that ultimately quench the free radicals generated by excess AIBN. These reactions may include radical addition to the OMRP dormant species R−CuII/TPMA*, to generate R−R coupling products, and/or addition to the H−CuII/TPMA* hydride intermediate of a CRT process to generate R−H, as shown in Scheme 3. PREDICI simulations confirmed the possibility of R−CuII/TPMA* or H−CuII/TPMA* formation, followed by a termination reaction with another radical to give CuI/TPMA* and the recombination product. Further investigations are planned to elucidate this mechanistic question. Nevertheless, in the presence of ATRP initiators, the ATRP equilibrium is primarily responsible for the control over the polymerization procedure and dominates over the OMRP processes.

S Supporting Information *

This material is available free of charge via the Internet at http://pubs.acs.org. 7998

dx.doi.org/10.1021/om3006883 | Organometallics 2012, 31, 7994−7999

Organometallics



Article

(d) Braunecker, W. A.; Tsarevsky, N. V.; Gennaro, A.; Matyjaszewski, K. Macromolecules 2009, 42, 6348. (e) Morick, J.; Buback, M.; Matyjaszewski, K. Macromol. Chem. Phys. 2011, 212, 2423. (16) Tang, W.; Kwak, Y.; Braunecker, W.; Tsarevsky, N. V.; Coote, M. L.; Matyjaszewski, K. J. Am. Chem. Soc. 2008, 130, 10702. (17) Magenau, A. J. D.; Kwak, Y.; Schröder, K.; Matyjaszewski, K. ACS Macro Lett. 2012, 508. (18) Schröder, K.; Mathers, R. T.; Buback, J.; Konkolewicz, D.; Magenau, A. J. D.; Matyjaszewski, K. ACS Macro Lett. 2012, 1, 1037. (19) (a) Le Grognec, E.; Claverie, J.; Poli, R. J. Am. Chem. Soc. 2001, 123, 9513. (b) Stoffelbach, F.; Poli, R.; Maria, S.; Richard, P. J. Organomet. Chem. 2007, 692, 3133. (20) Braunecker, W. A.; Itami, Y.; Matyjaszewski, K. Macromolecules 2005, 38, 9402. (21) Shaver, M. P.; Allan, L. E. N.; Gibson, V. C. Organometallics 2007, 26, 4725. (22) Allan, L. E. N.; MacDonald, J. P.; Reckling, A. M.; Kozak, C. M.; Shaver, M. P. Macromol. Rapid Commun. 2012, 33, 414. (23) Matyjaszewski, K.; Woodworth, B. E. Macromolecules 1998, 31, 4718. (24) (a) Kinoshita, I.; James Wright, L.; Kubo, S.; Kimura, K.; Sakata, A.; Yano, T.; Miyamoto, R.; Nishioka, T.; Isobe, K. Dalton Trans. 2003, 1993. (b) Goj, L. A.; Blue, E. D.; Delp, S. A.; Gunnoe, T. B.; Cundari, T. R.; Petersen, J. L. Organometallics 2006, 25, 4097. (25) Fischer, H.; Radom, L. Angew. Chem., Int. Ed. 2001, 40, 1340. (26) Tang, W.; Tsarevsky, N. V.; Matyjaszewski, K. J. Am. Chem. Soc. 2006, 128, 1598. (27) Jastrzebski, J. T. B. H.; van Koten, G. Structures and Reactivities of Organocopper Compounds in Modern Organocopper Chemistry; Krause, N., Ed.; Wiley-VCH: Weinheim, 2002. (28) (a) Navon, N.; Golub, G.; Cohen, H.; Meyerstein, D. Organometallics 1995, 14, 5670. (b) Navon, N.; Cohen, H.; Meyerstein, D. Inorg. Chem. 1997, 36, 3781. (29) Buback, M.; Kurz, C. H.; Schmaltz, C. Macromol. Chem. Phys. 1998, 199, 1721.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors would like to thank the CRP Consortium at Carnegie Mellon University and NSF (CHE-1026060, CHE1039870, CHE-0130903) for financial support and Dr. Johannes Buback for his help with the stopped-flow measurements. K.S. thanks the Deutsche Forschungsgemeinschaft (DFG) for her postdoctoral fellowship (SCHR 1314). R.P. thanks the Agence National de la Recherche (project OMRP, grant ANR 2010 BLANC 7101).



REFERENCES

(1) (a) Matyjaszewski, K.; Davis, T. P., Handbook of Radical Polymerization; Wiley-Interscience: Hoboken, 2002; p 936. (b) Moad, G. Radical Polymerization in Polymer Science: A Comprehensive Reference; Matyjaszewski, K.; Möller, M., Eds.; Elsevier: Amsterdam, 2012; Vol. 3, p 59. (2) Jenkins, A. D.; Jones, R. G.; Moad, G. Pur. Appl. Chem. 2010, 82, 483. (3) (a) Coessens, V.; Pintauer, T.; Matyjaszewski, K. Prog. Polym. Sci. 2001, 26, 337. (b) Matyjaszewski, K.; Davis, K. A. Statistical, Gradient and Segmented Copolymers by Controlled/Living Radical Polymerizations; Springer Verlag: Berlin, 2002; p 203. (c) Tsarevsky, N. V.; Matyjaszewski, K. Chem. Rev. 2007, 107, 2270. (d) Matyjaszewski, K.; Tsarevsky, N. V. Nat. Chem. 2009, 1, 276. (e) Wang, Y.; Matyjaszewski, K. Macromolecules 2011, 44, 1226. (f) Matyjaszewski, K. Macromolecules 2012, 45, 4015. (4) di Lena, F.; Matyjaszewski, K. Prog. Polym. Sci. 2010, 35, 959. (5) (a) Poli, R. Angew. Chem., Int. Ed. 2006, 45, 5058. (b) Poli, R. Organometallic Mediated Radical Polymerization in Polymer Science: A Comprehensive Reference; Matyjaszewski, K.; Möller, M., Eds.; Elsevier BV: Amsterdam, 2012; Vol. 3, p 351. (6) Wayland, B. B.; Poszmik, G.; Mukerjee, S. L.; Fryd, M. J. Am. Chem. Soc. 1994, 116, 7943. (7) (a) Smith, K. M.; McNeil, W. S.; Abd-El-Aziz, A. S. Macromol. Chem. .Phys. 2010, 211, 10. (b) Poli, R. Eur. J. Inorg. Chem. 2011, 2011, 1513. (c) Hurtgen, M.; Detrembleur, C.; Jerome, C.; Debuigne, A. Polym. Rev. 2011, 51, 188. (d) Allan, L. E. N.; Perry, M. R.; Shaver, M. P. Prog. Polym. Sci. 2012, 37, 127. (8) (a) Maria, S.; Kaneyoshi, H.; Matyjaszewski, K.; Poli, R. Chem. Eur. J. 2007, 13, 2480. (b) Santosh Kumar, K. S.; Gnanou, Y.; Champouret, Y.; Daran, J.-C.; Poli, R. Chem.−Eur. J. 2009, 15, 4874. (9) Zhao, Y.; Dong, H.; Li, Y.; Fu, X. Chem. Commun. 2012, 48, 3506. (10) Debuigne, A.; Morin, A. N.; Kermagoret, A.; Piette, Y.; Detrembleur, C.; Jérôme, C.; Poli, R. Chem.−Eur. J. 2012, 18, 12834− 12844. (11) Heuts, J. P. A.; Forster, D. J.; Davis, T. P.; Yamada, B.; Yamazoe, H.; Azukizawa, M. Macromolecules 1999, 32, 2511. (12) (a) Wang, J.-S.; Matyjaszewski, K. J. Am. Chem. Soc. 1995, 117, 5614. (b) Matyjaszewski, K.; Xia, J. Chem. Rev. 2001, 101, 2921. (c) Matyjaszewski, K.; Spanswick, J. In Polymer Science: A Comprehensive Reference; Matyjaszewski, K., Möller, M., Eds.; Elsevier B. V.: Amsterdam, The Netherlands, 2012; Vol. 3, p 377. (13) (a) Jakubowski, W.; Matyjaszewski, K. Angew. Chem., Int. Ed. 2006, 45, 4482. (b) Magenau, A. J. D.; Strandwitz, N. C.; Gennaro, A.; Matyjaszewski, K. Science 2011, 332, 81. (14) Matyjaszewski, K.; Jakubowski, W.; Min, K.; Tang, W.; Huang, J.; Braunecker, W. A.; Tsarevsky, N. V. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 15309. (15) (a) Tang, W.; Matyjaszewski, K. Macromolecules 2006, 39, 4953. (b) Tang, W.; Matyjaszewski, K. Macromolecules 2007, 40, 1858. (c) Seeliger, F.; Matyjaszewski, K. Macromolecules 2009, 42, 6050. 7999

dx.doi.org/10.1021/om3006883 | Organometallics 2012, 31, 7994−7999